-
Hi, It would be very helpful for those who are trying to train with new data and languages |
Beta Was this translation helpful? Give feedback.
Answered by
titu1994
Jun 8, 2023
Replies: 1 comment
-
It's not really tractable cause every used has different gpu with different memory. Providing A100 as a baseline is possible but there are too many variations of model configs to manually test the peak memory and batch size. |
Beta Was this translation helpful? Give feedback.
0 replies
Answer selected by
mehadi92
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
It's not really tractable cause every used has different gpu with different memory. Providing A100 as a baseline is possible but there are too many variations of model configs to manually test the peak memory and batch size.